Aim

We tested the hypothesis that exclusion from fire and cattle is responsible for the increase in tree cover in open savanna vegetation.

Location

Four plots in open savanna vegetation from the Calabozo region in central Venezuela were studied. Plot A was located in a Biological Station (EBL) that was excluded from fire and cattle between 1961 and 1991, with only two burning events in 1964 and 1968. The other plots (B, C, D) were located within 2 km distance from A, in neighbouring farms with soils similar to those in A but under various regimes of land use and fire frequency.

Methods

We measured the cover of isolated trees, small tree groups and groves of each plot in 1960 and 1977 using geographic information system (GIS) and digitalized aerial photographs. Additionally, the plots were located in the field and the open grassland was sampled in 1995 for species composition and density of stems above 20 cm height. Information on land use was obtained surveying people at the farms.

Results

There was an increase in the woody component of all plots during the 17-year interval (1960–1977). Total woody cover in the four plots as a whole increased from 4.5% to 17.9%. All three components measured, groves, tree groups and isolated trees, increased despite differences in land use and fire frequency between plots. Contrary to our expectations, the field survey performed in 1995 showed that fire-sensitive species were abundant in the open savanna in plots B, C and D, which were not excluded. Plot B, with the most intense agricultural use showed the highest rate of woody increase, and plot C, under extensive cattle ranching, was second. The results also showed that woody cover increased by aggregation from single trees and small tree groups into groves. As a consequence of these changes, savanna physiognomy changed from open to dense savanna parkland with a woody cover reaching over 25% in one of the four plots.

Conclusions

The results agree with other reported increases in woody cover in savannas under exclusion or with annual fires during the same time period in Africa (10). Our results support evidence from previous studies showing that fire and grazing are only part of a complex system of interacting factors affecting the structure of savanna communities.